Covid Cases Time Series [Example]

##### Installation and loading of required packages #####
if(!eval(parse(text="require(pacman)"))) 
{ 
  install.packages("pacman")
  eval(parse(text="require(pacman)"))
}

Lade n昼㸶tiges Paket: pacman

pacman::p_load(
  here,
  ggplot2,
  readr,
  dplyr,
  timetk,
  hrbrthemes,
  h2o,
  neptune
)

df <- read_csv(here::here("data/RKI_COVID19.csv"))

Parsed with column specification:
cols(
  ObjectId = col_double(),
  IdBundesland = col_double(),
  Bundesland = col_character(),
  Landkreis = col_character(),
  Altersgruppe = col_character(),
  Geschlecht = col_character(),
  AnzahlFall = col_double(),
  AnzahlTodesfall = col_double(),
  Meldedatum = col_character(),
  IdLandkreis = col_character(),
  Datenstand = col_character(),
  NeuerFall = col_double(),
  NeuerTodesfall = col_double(),
  Refdatum = col_character(),
  NeuGenesen = col_double(),
  AnzahlGenesen = col_double(),
  IstErkrankungsbeginn = col_double(),
  Altersgruppe2 = col_character()
)

df$Meldedatum <- as.Date(df$Meldedatum)

filterLandkreis <- "SK Münster"
df_m <- df %>% filter(Landkreis == filterLandkreis)
df_agg <- df_m %>% group_by (Meldedatum) %>% summarize (faelle = sum(AnzahlFall))

`summarise()` ungrouping output (override with `.groups` argument)

myTheme <- theme_ft_rc()
theme_set(myTheme)

 df_agg %>% timetk::plot_anomaly_diagnostics(Meldedatum, faelle)

frequency = 6 observations per 1 week
trend = 75 observations per 3 months
Registered S3 method overwritten by 'htmlwidgets':
  method           from         
  print.htmlwidget tools:rstudio

 ts <- df_agg %>% tk_augment_slidify(.value  = faelle, .period  = c(2, 6, 30), .f = mean, .partial = TRUE)

New names:
* NA -> ...1
* NA -> ...2
* NA -> ...3

 h2o.init()

H2O is not running yet, starting it now...

Note:  In case of errors look at the following log files:
    C:\Users\kof\AppData\Local\Temp\RtmpOOO9qu\file445043515609/h2o_KoF_started_from_r.out
    C:\Users\kof\AppData\Local\Temp\RtmpOOO9qu\file44504dd84bfe/h2o_KoF_started_from_r.err

java version "11.0.2" 2019-01-15 LTS
Java(TM) SE Runtime Environment 18.9 (build 11.0.2+9-LTS)
Java HotSpot(TM) 64-Bit Server VM 18.9 (build 11.0.2+9-LTS, mixed mode)

Starting H2O JVM and connecting: . Connection successful!

R is connected to the H2O cluster: 
    H2O cluster uptime:         9 seconds 82 milliseconds 
    H2O cluster timezone:       Europe/Berlin 
    H2O data parsing timezone:  UTC 
    H2O cluster version:        3.32.0.1 
    H2O cluster version age:    2 months and 27 days  
    H2O cluster name:           H2O_started_from_R_KoF_hak113 
    H2O cluster total nodes:    1 
    H2O cluster total memory:   3.97 GB 
    H2O cluster total cores:    8 
    H2O cluster allowed cores:  8 
    H2O cluster healthy:        TRUE 
    H2O Connection ip:          localhost 
    H2O Connection port:        54321 
    H2O Connection proxy:       NA 
    H2O Internal Security:      FALSE 
    H2O API Extensions:         Amazon S3, Algos, AutoML, Core V3, TargetEncoder, Core V4 
    R Version:                  R version 3.6.3 (2020-02-29) 

 df_h2o = as.h2o(ts)

Registered S3 method overwritten by 'data.table':
  method           from
  print.data.table     

  |                                                                                                                                                  
  |                                                                                                                                            |   0%
  |                                                                                                                                                  
  |============================================================================================================================================| 100%

 model <- h2o.automl(training_frame = df_h2o, 
            nfolds = 3,
            max_runtime_secs = 5,
            x = c("faelle_roll_2","faelle_roll_6","faelle_roll_30"),
            y = "faelle", exclude_algos = "StackedEnsemble")

  |                                                                                                                                                  
  |                                                                                                                                            |   0%
  |                                                                                                                                                  
  |=====                                                                                                                                       |   4%
08:56:39.904: AutoML: XGBoost is not available; skipping it.
  |                                                                                                                                                  
  |=============================================================================                                                               |  55%
  |                                                                                                                                                  
  |======================================================================================================================================      |  96%
  |                                                                                                                                                  
  |============================================================================================================================================| 100%

 pred <- cbind(ts, as.data.frame(h2o.predict(model, df_h2o)))

  |                                                                                                                                                  
  |                                                                                                                                            |   0%
  |                                                                                                                                                  
  |============================================================================================================================================| 100%

init_neptune(project_name = "frank.koehne/covid-drift",
    api_token = Sys.getenv("NEPTUNE_API_TOKEN")
)

WARNING: There is a new version of neptune-client 0.4.130 (installed: 0.4.129).

Project(frank.koehne/covid-drift)

create_experiment(name = "Rolling Avg",
    tags = c("experimental", "automl", "roll2", "roll6", "roll30"),
    params = list(tuneLength = 5, model = "automl")
)

https://ui.neptune.ai/frank.koehne/covid-drift/e/COV-4
Experiment(COV-4)

set_property(property = "data-version",
    value = max(df_m$Datenstand)
)

log_metric("n", nrow(df_agg))
log_metric("X-Val-RMSE", h2o.rmse(model@leader, xval = TRUE))

Results can be seen publicly.